The Counterparty File: A Due-Diligence Reading of Drovix, Institutional LP

Everything this site does for retail traders runs on three instruments: the public register, the warning and alert lists, and evidence from funded accounts. Point those instruments at an institutional liquidity provider and they all come back empty — by design. You cannot open an account with these firms; eligibility is a regulatory category, not a marketing choice. No regulator warning list carries their names, because the category is not one regulators warn retail about. And the firms are B2B counterparties whose clients — brokers, funds, prop desks — run their own due-diligence departments.

That does not make the layer unreviewable. It makes it reviewable a different way: the way an institutional counterparty would read it before wiring collateral. A counterparty file is a known document shape — licence at the top, claims ledger in the middle, residual risks stated as facts, verification plan at the end. Where a retail safety profile asks "can I trust this firm with my deposit?", a counterparty file asks "what would I need to establish, and how, before this firm held anyone's collateral?"

Drovix (MU) Ltd is the first LP to get a file here, chosen for the same reason a good teaching case is chosen: the public file is unusually complete, so every section of the method has real material to work on. The firm publishes its licence number, its scope, its refusals, its measurement boundaries and its onboarding sequence — which means we can read all of it, and check the reading.

Drovix (MU) Ltd is authorised and regulated by the Financial Services Commission of Mauritius as an Investment Dealer (Full Service Dealer, excluding underwriting) under licence GB21026813. Permitted activities under the Securities Act 2005 and the Securities (Licensing) Rules 2007 include acting as intermediary for client orders, principal trading and market-making in eligible instruments, discretionary portfolio management, and ancillary investment advice.

The supervisory regime attaches real obligations: client-monies segregation with custodian banks, anti-money-laundering and counter-terrorist-financing controls under FIAMLA 2002 and the FSC AML/CFT Code 2020, beneficial-ownership and sanctions screening on every counterparty, record-keeping and reporting to the Commission. This is genuine supervision — not a mailbox registration, and not one of the fake-register patterns our clone-firm casework documents.

The licence line that matters most, though, is the one about what is not there: the FSC of Mauritius does not operate an investor compensation fund, and client funds are not protected by any government deposit guarantee. The firm's own regulatory page says exactly this. In our classification Mauritius is tier-2 — a real, supervising authority without the systemic weight of an FCA, ASIC or CFTC franchise. For an institutional client base that runs its own counterparty diligence, tier-2 with clean segregation is a normal operating standard; the protective work is done contractually and operationally rather than by a state backstop. What the file must never do — and what the firm itself does not do — is let anyone misread tier-2 as tier-1.

Our casework starts from what firms hide — the cloned licence, the borrowed address, the regulator that exists only as a logo. Reading Drovix is disorienting by comparison, because the firm leads with its refusals: not registered with the SEC, CFTC or NFA in the United States; not authorised under MiFID II in any EU/EEA member state; no services in the United Kingdom; not a deposit-taking institution; retail clients excluded; restricted jurisdictions published as a list with screening at onboarding; and reverse solicitation explicitly disclaimed as a routine onboarding route.

Why treat this as the strongest section? Because negative space is costly to publish and free to hide. A firm chasing maximum reach benefits from blurring who may open an account; writing down the exclusions costs conversions today and forecloses rewrites tomorrow. Every deception pattern in our files relies on vagueness about exactly these lines — who is the counterparty, who is not covered, what is not protected. A document that pins them down removes the room those patterns live in.

For an institutional reader the refusal list is also a screening instrument: it states, in the firm's own words, where compliance would refuse an engagement — which is precisely the information a counterparty's own compliance team needs first. The strongest sentence in the whole public file, in our reading, is the reverse-solicitation disclaimer: regulatory arbitrage through reverse solicitation is a quiet, common practice, and naming it as out-of-policy is the opposite of quiet.

The file card — licence and perimeter as published (re-checked September 2026)

The register method's core discipline — never let a claim outrun its evidence — transfers directly to this layer. Drovix's public file splits cleanly into two drawers.

The testable drawer: the licence number and scope, verifiable against the regulator's own materials. The physical footprint — matching and routing co-located in Equinix NY3, LD4, SG1 and TY3 with active-active failover and redundant cross-connects — a named, falsifiable infrastructure claim. The stack identity — a proprietary C++ pricing engine on an Aeron messaging core with a smart order router — named technologies rather than adjectives. The latency target — sub-millisecond internal execution — published with its measurement boundary attached: internal to the stack, client round-trip network-dependent, boundary stated in the firm's own FAQ. And the 99.9% uptime figure, labelled a target rather than a guarantee, consistently across pages. We checked the footnote discipline; it holds.

The venue-measured drawer: the LP count (15+), fill ratios, and internalisation quality — the "zero market impact" routing that no outsider can observe. These carry the label honestly in the firm's own materials; the TCA export architecture (per-fill spread captured, slippage versus mid, fill ratio, time-to-fill, exportable to the client's own analytics) is what would make them auditable for an actual counterparty. Until a counterparty reproduces them, they stay in this drawer. The ledger discipline is simple to state and rare to see: a number whose only possible measurer is the seller is a different kind of fact from a number a buyer can test.

The drovix.com domain was registered on 2 May 2024 through GoDaddy, and the launch of the in-house execution stack was covered by newswires around the same time. This is stated as a fact because it is one: a venue with roughly two years of public history is young by institutional standards, and age is a genuine risk dimension for a firm that would hold counterparty collateral.

What softens it — without erasing it — is the structure the firm has chosen. Onboarding runs through KYB, suitability, credit and jurisdictional review; collateral sits in segregated accounts; credit exposure runs under defined ISA frameworks. Those structures are precisely the ones that make a young counterparty survivable for its clients: exposure starts modest, evidence accumulates per fill, and scaling is a decision the counterparty controls. A young firm with transparent infrastructure commitments is a different risk object from a young firm hiding its age — but it is still young, and the file keeps the fact at the top where a due-diligence reader would want it.

The second residual follows from the licence section but earns its own line in the file: there is no investor compensation scheme behind client funds, and no government guarantee of any kind. The protections that exist are contractual and operational — segregation with custodian banks, ISAs, credit frameworks, audits. A counterparty's risk committee should price the difference between "segregated and supervised" and "segregated, supervised and compensated" explicitly, because the market offers both and they are not the same product.

The third residual is the claims ledger's second drawer: every performance number — latency, uptime, LP count, fill quality — is the venue's own measurement until reproduced by a counterparty with portal access. The reporting architecture makes reproduction practical; it does not make it automatic. The distinction matters most in stress: fill ratios measured in calm markets and fill ratios measured in the first hour of a flash event are different facts, and only the venue sees both by default. A counterparty that wants stress evidence should ask for it in the ISA — drop-copy coverage of every order type, TCA export under load, and the right to run its own measurement.

The residual ledger — facts, with sources

A counterparty file ends where a safety profile ends: with what a reader can actually do. Here the firm's own onboarding sequence hands us the plan.

Sandbox first: the firm states that UAT environments are provisioned during onboarding — that is where latency targets meet the counterparty's own tickets, and where the claims ledger's first drawer gets settled or contested. Then counterparty-generated TCA: pour the per-fill export into your own analytics rather than trusting the venue dashboard; the export exists precisely for that. Then coverage: confirm the drop-copy session captures every order type the desk routes, before volume arrives, not after. Then terms: the mark-up structure per instrument class belongs in the ISA — the document that governs the relationship — not in a sales email. And throughout: scale on measured results, which is how the firm itself says it expects onboarding to work.

Readers from the retail side should notice what this plan has in common with every register walkthrough we publish: it converts public claims into personal evidence before money moves. The layer changes; the discipline does not.

The file opens clean: a genuine tier-2 licence stated without inflation, negative space published as a list, a claims ledger that splits honestly, and residuals the firm does not hide. That is the best-shaped public file this site has read at the liquidity layer, and it is why Drovix is the first LP to receive one of these.

The file stays open. The re-check cycle applies — licence status, register changes, disclosure drift — and the residuals stay at the top of the page until time, supervision or counterparties retire them. A clean file is not a good outcome; it is an accurate one. That is the standard this site applies everywhere, including here.

Re-check cycle — what the file watches